COMPUTATIONAL SOCIAL SCIENCE

CSS SEMINAR - February 12 - TBA

Friday, February 12, 3:00 p.m.
Center for Social Complexity Suite
3rd Floor, Research Hall

Dale Brearcliffe, Peter Froncek, Marta Hansen, Vince Kane, Stefan McCabe, Davoud Taghawi-Nejad and Rob Axtell

A Comparison of Languages and Frameworks for the Parallelization of a Simple Agent Model

ABSTRACT: The zero intelligence trader model is a well-known agent system representing supply and demand relations and the emergence of market clearing. We have parallelized it using a variety of languages and technologies, including pthreads in C, C11 threads, Clojure, Erlang, Go, Haskell, Java, Python, and Scala. For each, significant speed-ups as a function of the number of threads is demonstrated. The advantages and disadvantages of each approach are described and compared. Some of these languages are largely functional, some feature persistent data, and some have relatively poor absolute performance. We discuss the utility of them for multi-agent systems generally, including ease of programming and likely performance gains.